23 Jan 2009 | Massimiliano Di Ventra, Yuriy V. Pershin, and Leon O. Chua, Fellow, IEEE
This paper introduces memristors, memcapacitors, and meminductors—circuit elements with memory. These devices, like memristors, have properties that depend on the system's state and history. They exhibit pinched hysteretic loops in their constitutive variables: current-voltage for memristors, charge-voltage for memcapacitors, and current-flux for meminductors. These elements are common at the nanoscale where electron and ion dynamics depend on the system's history. They can be used in electronics for new functionalities, especially in neuromorphic devices to simulate learning, adaptive, and spontaneous behavior.
Memristors, first proposed by Chua, have a resistance that depends on the system's internal state. Memcapacitors and meminductors generalize this concept to capacitive and inductive systems. These devices can store energy, unlike memristors. Memcapacitors have a capacitance that depends on the system's state, while meminductors have an inductance that depends on the system's state.
Memristive systems are passive, cannot store energy, and exhibit a "pinched hysteretic loop" in their voltage-current characteristics. They behave as linear resistors at high frequencies and nonlinear resistors at low frequencies. Memcapacitive systems behave as linear capacitors at high frequencies and nonlinear capacitors at low frequencies. Meminductive systems behave as linear inductors at high frequencies and nonlinear inductors at low frequencies.
Examples of memristive systems include thermistors, molecular systems, spintronic devices, and thin film nanostructures. Memcapacitive systems include nanoscale capacitors with interface traps or embedded nanocrystals. Meminductive systems include inductors with core materials whose response depends on their history.
These devices have potential applications in neuromorphic devices, non-volatile memories, and other electronic systems. They can simulate learning, adaptive, and spontaneous behavior. The paper concludes that these devices are versatile and their combined operations in electronic circuits are still largely unexplored.This paper introduces memristors, memcapacitors, and meminductors—circuit elements with memory. These devices, like memristors, have properties that depend on the system's state and history. They exhibit pinched hysteretic loops in their constitutive variables: current-voltage for memristors, charge-voltage for memcapacitors, and current-flux for meminductors. These elements are common at the nanoscale where electron and ion dynamics depend on the system's history. They can be used in electronics for new functionalities, especially in neuromorphic devices to simulate learning, adaptive, and spontaneous behavior.
Memristors, first proposed by Chua, have a resistance that depends on the system's internal state. Memcapacitors and meminductors generalize this concept to capacitive and inductive systems. These devices can store energy, unlike memristors. Memcapacitors have a capacitance that depends on the system's state, while meminductors have an inductance that depends on the system's state.
Memristive systems are passive, cannot store energy, and exhibit a "pinched hysteretic loop" in their voltage-current characteristics. They behave as linear resistors at high frequencies and nonlinear resistors at low frequencies. Memcapacitive systems behave as linear capacitors at high frequencies and nonlinear capacitors at low frequencies. Meminductive systems behave as linear inductors at high frequencies and nonlinear inductors at low frequencies.
Examples of memristive systems include thermistors, molecular systems, spintronic devices, and thin film nanostructures. Memcapacitive systems include nanoscale capacitors with interface traps or embedded nanocrystals. Meminductive systems include inductors with core materials whose response depends on their history.
These devices have potential applications in neuromorphic devices, non-volatile memories, and other electronic systems. They can simulate learning, adaptive, and spontaneous behavior. The paper concludes that these devices are versatile and their combined operations in electronic circuits are still largely unexplored.